@ax-llm/ax@25.3.2
The best library to work with LLMs
AI security review
scanned · by lpm-firewall-aiThe verified lifecycle hook performs automatic, project-local setup of the package’s own Claude skills, which warrants a warning for agent extension lifecycle risk. The non-UTF-8 runtime entrypoints remain unresolved, so the source review does not support a clean verdict; no inspected behavior establishes a concrete attack for blocking.